T-Norm of Yager Class of Subsethood Defuzzification: Improving Enrolment Forecast in Fuzzy Time Series
Abstract
Fuzzy time series has been used to model observations that contain multiple values. This paper proposes the t-norm of Yager class of subsethood defuzzification to forecast university enrolments based on fuzzy time series and the data of historical enrolments which are adopted from Song and Chissom (1994). The proposed method applied seven and ten interval with equal length and the max-product and max-min as the composition operator in the fuzzy relations F(t)= F(t −1) R(t, t −1). The result shows that the t-norm of Yager class of subsethood defuzzification models with (10, max-product) is the best forecasting method in terms of accuracy. The proposed method has also improved the forecasting results by previous researchers.
Copyright (c) 2008 admin

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
Authors who publish with this journal agree to the following terms:
1. Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
2. Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
3. Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work.